Upload 9 files
Browse files- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +30 -0
- tokenizer.json +0 -0
- tokenizer.model +3 -0
- tokenizer_config.json +42 -0
- trainer_state.json +751 -0
- training_args.bin +3 -0
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:92c36a898cd8e2dd7bbc58deb64c721813aba6d5fc055ec40dfabed08d43244f
|
| 3 |
+
size 36072070
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:527e76589de1ce09084875a615d5046549d45eccd68a6a1373f7a1358349cc7c
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fbbfa411208c0097956115f6349b400eeedf28dc4897f54f081a71275edc25b8
|
| 3 |
+
size 1064
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<unk>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"0": {
|
| 6 |
+
"content": "<unk>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"1": {
|
| 14 |
+
"content": "<s>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"2": {
|
| 22 |
+
"content": "</s>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
}
|
| 29 |
+
},
|
| 30 |
+
"bos_token": "<s>",
|
| 31 |
+
"chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
|
| 32 |
+
"clean_up_tokenization_spaces": false,
|
| 33 |
+
"eos_token": "</s>",
|
| 34 |
+
"legacy": false,
|
| 35 |
+
"model_max_length": 2048,
|
| 36 |
+
"pad_token": "</s>",
|
| 37 |
+
"padding_side": "right",
|
| 38 |
+
"sp_model_kwargs": {},
|
| 39 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 40 |
+
"unk_token": "<unk>",
|
| 41 |
+
"use_default_system_prompt": false
|
| 42 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,751 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"learning_rate": 6.6666666666666675e-06,
|
| 14 |
+
"logits/chosen": -3.1526219844818115,
|
| 15 |
+
"logits/rejected": -3.3119924068450928,
|
| 16 |
+
"logps/chosen": -18.28135108947754,
|
| 17 |
+
"logps/rejected": -33.52398681640625,
|
| 18 |
+
"loss": 0.6997,
|
| 19 |
+
"rewards/accuracies": 0.0,
|
| 20 |
+
"rewards/chosen": -0.006796836853027344,
|
| 21 |
+
"rewards/margins": -0.012901116162538528,
|
| 22 |
+
"rewards/rejected": 0.006104278843849897,
|
| 23 |
+
"step": 1
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04,
|
| 27 |
+
"learning_rate": 6.666666666666667e-05,
|
| 28 |
+
"logits/chosen": -3.1073851585388184,
|
| 29 |
+
"logits/rejected": -3.090308666229248,
|
| 30 |
+
"logps/chosen": -20.141502380371094,
|
| 31 |
+
"logps/rejected": -18.037580490112305,
|
| 32 |
+
"loss": 0.694,
|
| 33 |
+
"rewards/accuracies": 0.0833333358168602,
|
| 34 |
+
"rewards/chosen": 0.001263597165234387,
|
| 35 |
+
"rewards/margins": -0.001611270010471344,
|
| 36 |
+
"rewards/rejected": 0.002874867059290409,
|
| 37 |
+
"step": 10
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08,
|
| 41 |
+
"learning_rate": 0.00013333333333333334,
|
| 42 |
+
"logits/chosen": -3.0630269050598145,
|
| 43 |
+
"logits/rejected": -3.1416983604431152,
|
| 44 |
+
"logps/chosen": -31.93638038635254,
|
| 45 |
+
"logps/rejected": -42.507789611816406,
|
| 46 |
+
"loss": 0.6916,
|
| 47 |
+
"rewards/accuracies": 0.2750000059604645,
|
| 48 |
+
"rewards/chosen": 0.010843334719538689,
|
| 49 |
+
"rewards/margins": 0.003288193140178919,
|
| 50 |
+
"rewards/rejected": 0.0075551411136984825,
|
| 51 |
+
"step": 20
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12,
|
| 55 |
+
"learning_rate": 0.0002,
|
| 56 |
+
"logits/chosen": -3.148637294769287,
|
| 57 |
+
"logits/rejected": -3.150296211242676,
|
| 58 |
+
"logps/chosen": -22.95195770263672,
|
| 59 |
+
"logps/rejected": -23.612133026123047,
|
| 60 |
+
"loss": 0.6974,
|
| 61 |
+
"rewards/accuracies": 0.15000000596046448,
|
| 62 |
+
"rewards/chosen": -0.002520971465855837,
|
| 63 |
+
"rewards/margins": -0.0078018950298428535,
|
| 64 |
+
"rewards/rejected": 0.005280924029648304,
|
| 65 |
+
"step": 30
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.16,
|
| 69 |
+
"learning_rate": 0.0002666666666666667,
|
| 70 |
+
"logits/chosen": -3.1271812915802,
|
| 71 |
+
"logits/rejected": -3.0903429985046387,
|
| 72 |
+
"logps/chosen": -34.57979965209961,
|
| 73 |
+
"logps/rejected": -27.37040138244629,
|
| 74 |
+
"loss": 0.6925,
|
| 75 |
+
"rewards/accuracies": 0.125,
|
| 76 |
+
"rewards/chosen": 0.03781440109014511,
|
| 77 |
+
"rewards/margins": 0.006826506461948156,
|
| 78 |
+
"rewards/rejected": 0.030987894162535667,
|
| 79 |
+
"step": 40
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.2,
|
| 83 |
+
"learning_rate": 0.0003333333333333333,
|
| 84 |
+
"logits/chosen": -3.0824079513549805,
|
| 85 |
+
"logits/rejected": -3.0999526977539062,
|
| 86 |
+
"logps/chosen": -29.16314697265625,
|
| 87 |
+
"logps/rejected": -30.843231201171875,
|
| 88 |
+
"loss": 0.7057,
|
| 89 |
+
"rewards/accuracies": 0.25,
|
| 90 |
+
"rewards/chosen": 0.08534860610961914,
|
| 91 |
+
"rewards/margins": -0.013902002945542336,
|
| 92 |
+
"rewards/rejected": 0.09925060719251633,
|
| 93 |
+
"step": 50
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.24,
|
| 97 |
+
"learning_rate": 0.0004,
|
| 98 |
+
"logits/chosen": -3.1495282649993896,
|
| 99 |
+
"logits/rejected": -3.184638500213623,
|
| 100 |
+
"logps/chosen": -19.72355079650879,
|
| 101 |
+
"logps/rejected": -28.213886260986328,
|
| 102 |
+
"loss": 0.7213,
|
| 103 |
+
"rewards/accuracies": 0.07500000298023224,
|
| 104 |
+
"rewards/chosen": -0.05634387582540512,
|
| 105 |
+
"rewards/margins": -0.043291497975587845,
|
| 106 |
+
"rewards/rejected": -0.01305237878113985,
|
| 107 |
+
"step": 60
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.28,
|
| 111 |
+
"learning_rate": 0.00046666666666666666,
|
| 112 |
+
"logits/chosen": -3.04826021194458,
|
| 113 |
+
"logits/rejected": -3.005667209625244,
|
| 114 |
+
"logps/chosen": -25.625635147094727,
|
| 115 |
+
"logps/rejected": -24.768199920654297,
|
| 116 |
+
"loss": 0.7002,
|
| 117 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 118 |
+
"rewards/chosen": -0.06301303952932358,
|
| 119 |
+
"rewards/margins": -0.00712633365765214,
|
| 120 |
+
"rewards/rejected": -0.055886708199977875,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.32,
|
| 125 |
+
"learning_rate": 0.0005333333333333334,
|
| 126 |
+
"logits/chosen": -3.048518419265747,
|
| 127 |
+
"logits/rejected": -3.061340808868408,
|
| 128 |
+
"logps/chosen": -32.414894104003906,
|
| 129 |
+
"logps/rejected": -34.773399353027344,
|
| 130 |
+
"loss": 0.7835,
|
| 131 |
+
"rewards/accuracies": 0.32499998807907104,
|
| 132 |
+
"rewards/chosen": -0.23218408226966858,
|
| 133 |
+
"rewards/margins": -0.03052915260195732,
|
| 134 |
+
"rewards/rejected": -0.20165491104125977,
|
| 135 |
+
"step": 80
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.36,
|
| 139 |
+
"learning_rate": 0.0006,
|
| 140 |
+
"logits/chosen": -3.035808563232422,
|
| 141 |
+
"logits/rejected": -3.1053929328918457,
|
| 142 |
+
"logps/chosen": -31.509021759033203,
|
| 143 |
+
"logps/rejected": -45.353553771972656,
|
| 144 |
+
"loss": 0.7051,
|
| 145 |
+
"rewards/accuracies": 0.30000001192092896,
|
| 146 |
+
"rewards/chosen": -0.3204951286315918,
|
| 147 |
+
"rewards/margins": 0.283873975276947,
|
| 148 |
+
"rewards/rejected": -0.604369044303894,
|
| 149 |
+
"step": 90
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.4,
|
| 153 |
+
"learning_rate": 0.0006666666666666666,
|
| 154 |
+
"logits/chosen": -2.959228992462158,
|
| 155 |
+
"logits/rejected": -2.98266339302063,
|
| 156 |
+
"logps/chosen": -51.024803161621094,
|
| 157 |
+
"logps/rejected": -54.91625213623047,
|
| 158 |
+
"loss": 1.0868,
|
| 159 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 160 |
+
"rewards/chosen": -1.835883378982544,
|
| 161 |
+
"rewards/margins": 0.3176426589488983,
|
| 162 |
+
"rewards/rejected": -2.1535260677337646,
|
| 163 |
+
"step": 100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.44,
|
| 167 |
+
"learning_rate": 0.0007333333333333333,
|
| 168 |
+
"logits/chosen": -3.0634117126464844,
|
| 169 |
+
"logits/rejected": -3.0850846767425537,
|
| 170 |
+
"logps/chosen": -23.597423553466797,
|
| 171 |
+
"logps/rejected": -26.58676528930664,
|
| 172 |
+
"loss": 0.7629,
|
| 173 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 174 |
+
"rewards/chosen": -0.2058214694261551,
|
| 175 |
+
"rewards/margins": -0.05166854336857796,
|
| 176 |
+
"rewards/rejected": -0.15415294468402863,
|
| 177 |
+
"step": 110
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.48,
|
| 181 |
+
"learning_rate": 0.0008,
|
| 182 |
+
"logits/chosen": -3.1232190132141113,
|
| 183 |
+
"logits/rejected": -3.1285691261291504,
|
| 184 |
+
"logps/chosen": -20.47592544555664,
|
| 185 |
+
"logps/rejected": -22.593311309814453,
|
| 186 |
+
"loss": 0.7341,
|
| 187 |
+
"rewards/accuracies": 0.15000000596046448,
|
| 188 |
+
"rewards/chosen": -0.46454209089279175,
|
| 189 |
+
"rewards/margins": -0.023245975375175476,
|
| 190 |
+
"rewards/rejected": -0.44129619002342224,
|
| 191 |
+
"step": 120
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.52,
|
| 195 |
+
"learning_rate": 0.0008666666666666667,
|
| 196 |
+
"logits/chosen": -2.975956678390503,
|
| 197 |
+
"logits/rejected": -3.027247190475464,
|
| 198 |
+
"logps/chosen": -34.07396697998047,
|
| 199 |
+
"logps/rejected": -42.46125793457031,
|
| 200 |
+
"loss": 0.723,
|
| 201 |
+
"rewards/accuracies": 0.2750000059604645,
|
| 202 |
+
"rewards/chosen": -0.4881093502044678,
|
| 203 |
+
"rewards/margins": 0.3038038909435272,
|
| 204 |
+
"rewards/rejected": -0.7919132113456726,
|
| 205 |
+
"step": 130
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.56,
|
| 209 |
+
"learning_rate": 0.0009333333333333333,
|
| 210 |
+
"logits/chosen": -2.981985569000244,
|
| 211 |
+
"logits/rejected": -2.9559457302093506,
|
| 212 |
+
"logps/chosen": -33.598899841308594,
|
| 213 |
+
"logps/rejected": -40.866451263427734,
|
| 214 |
+
"loss": 0.7877,
|
| 215 |
+
"rewards/accuracies": 0.2750000059604645,
|
| 216 |
+
"rewards/chosen": -0.35516494512557983,
|
| 217 |
+
"rewards/margins": 0.4116950035095215,
|
| 218 |
+
"rewards/rejected": -0.7668598890304565,
|
| 219 |
+
"step": 140
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.6,
|
| 223 |
+
"learning_rate": 0.001,
|
| 224 |
+
"logits/chosen": -3.05903959274292,
|
| 225 |
+
"logits/rejected": -3.0611279010772705,
|
| 226 |
+
"logps/chosen": -20.234691619873047,
|
| 227 |
+
"logps/rejected": -19.169904708862305,
|
| 228 |
+
"loss": 0.9163,
|
| 229 |
+
"rewards/accuracies": 0.10000000149011612,
|
| 230 |
+
"rewards/chosen": -0.36899399757385254,
|
| 231 |
+
"rewards/margins": -0.08086968958377838,
|
| 232 |
+
"rewards/rejected": -0.28812432289123535,
|
| 233 |
+
"step": 150
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.64,
|
| 237 |
+
"learning_rate": 0.0009723756906077348,
|
| 238 |
+
"logits/chosen": -2.9612436294555664,
|
| 239 |
+
"logits/rejected": -2.9223554134368896,
|
| 240 |
+
"logps/chosen": -32.5883674621582,
|
| 241 |
+
"logps/rejected": -25.370834350585938,
|
| 242 |
+
"loss": 0.7856,
|
| 243 |
+
"rewards/accuracies": 0.17499999701976776,
|
| 244 |
+
"rewards/chosen": -0.18723489344120026,
|
| 245 |
+
"rewards/margins": -0.10649768263101578,
|
| 246 |
+
"rewards/rejected": -0.08073721826076508,
|
| 247 |
+
"step": 160
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.68,
|
| 251 |
+
"learning_rate": 0.0009447513812154696,
|
| 252 |
+
"logits/chosen": -2.9623398780822754,
|
| 253 |
+
"logits/rejected": -2.914522171020508,
|
| 254 |
+
"logps/chosen": -39.70682907104492,
|
| 255 |
+
"logps/rejected": -33.20659637451172,
|
| 256 |
+
"loss": 0.898,
|
| 257 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 258 |
+
"rewards/chosen": -0.4028944969177246,
|
| 259 |
+
"rewards/margins": -0.04350559413433075,
|
| 260 |
+
"rewards/rejected": -0.35938888788223267,
|
| 261 |
+
"step": 170
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.72,
|
| 265 |
+
"learning_rate": 0.0009171270718232044,
|
| 266 |
+
"logits/chosen": -3.0877394676208496,
|
| 267 |
+
"logits/rejected": -3.092603921890259,
|
| 268 |
+
"logps/chosen": -43.16739273071289,
|
| 269 |
+
"logps/rejected": -47.593746185302734,
|
| 270 |
+
"loss": 1.2201,
|
| 271 |
+
"rewards/accuracies": 0.17499999701976776,
|
| 272 |
+
"rewards/chosen": -2.5319814682006836,
|
| 273 |
+
"rewards/margins": -0.10082467645406723,
|
| 274 |
+
"rewards/rejected": -2.431157112121582,
|
| 275 |
+
"step": 180
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.76,
|
| 279 |
+
"learning_rate": 0.0008895027624309392,
|
| 280 |
+
"logits/chosen": -2.09025239944458,
|
| 281 |
+
"logits/rejected": -2.0903568267822266,
|
| 282 |
+
"logps/chosen": -112.7010269165039,
|
| 283 |
+
"logps/rejected": -105.31596374511719,
|
| 284 |
+
"loss": 2.9619,
|
| 285 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 286 |
+
"rewards/chosen": -7.883659362792969,
|
| 287 |
+
"rewards/margins": -0.7001466751098633,
|
| 288 |
+
"rewards/rejected": -7.1835126876831055,
|
| 289 |
+
"step": 190
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.8,
|
| 293 |
+
"learning_rate": 0.0008618784530386741,
|
| 294 |
+
"logits/chosen": -1.8135408163070679,
|
| 295 |
+
"logits/rejected": -1.8160464763641357,
|
| 296 |
+
"logps/chosen": -75.82559967041016,
|
| 297 |
+
"logps/rejected": -65.90226745605469,
|
| 298 |
+
"loss": 2.555,
|
| 299 |
+
"rewards/accuracies": 0.17499999701976776,
|
| 300 |
+
"rewards/chosen": -5.0599846839904785,
|
| 301 |
+
"rewards/margins": -0.6209059953689575,
|
| 302 |
+
"rewards/rejected": -4.439078330993652,
|
| 303 |
+
"step": 200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.84,
|
| 307 |
+
"learning_rate": 0.0008342541436464089,
|
| 308 |
+
"logits/chosen": -2.6365649700164795,
|
| 309 |
+
"logits/rejected": -2.633017063140869,
|
| 310 |
+
"logps/chosen": -73.8572998046875,
|
| 311 |
+
"logps/rejected": -115.65068054199219,
|
| 312 |
+
"loss": 1.3204,
|
| 313 |
+
"rewards/accuracies": 0.3499999940395355,
|
| 314 |
+
"rewards/chosen": -4.614927291870117,
|
| 315 |
+
"rewards/margins": 3.1844677925109863,
|
| 316 |
+
"rewards/rejected": -7.799394130706787,
|
| 317 |
+
"step": 210
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.88,
|
| 321 |
+
"learning_rate": 0.0008066298342541437,
|
| 322 |
+
"logits/chosen": -2.338550329208374,
|
| 323 |
+
"logits/rejected": -2.337129831314087,
|
| 324 |
+
"logps/chosen": -75.24410247802734,
|
| 325 |
+
"logps/rejected": -89.40665435791016,
|
| 326 |
+
"loss": 1.6189,
|
| 327 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 328 |
+
"rewards/chosen": -5.084308624267578,
|
| 329 |
+
"rewards/margins": 0.9201302528381348,
|
| 330 |
+
"rewards/rejected": -6.004438877105713,
|
| 331 |
+
"step": 220
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.92,
|
| 335 |
+
"learning_rate": 0.0007790055248618785,
|
| 336 |
+
"logits/chosen": -2.328997850418091,
|
| 337 |
+
"logits/rejected": -2.326946496963501,
|
| 338 |
+
"logps/chosen": -62.543540954589844,
|
| 339 |
+
"logps/rejected": -84.85903930664062,
|
| 340 |
+
"loss": 1.3971,
|
| 341 |
+
"rewards/accuracies": 0.22499999403953552,
|
| 342 |
+
"rewards/chosen": -3.8174233436584473,
|
| 343 |
+
"rewards/margins": 1.793341875076294,
|
| 344 |
+
"rewards/rejected": -5.610764503479004,
|
| 345 |
+
"step": 230
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.96,
|
| 349 |
+
"learning_rate": 0.0007513812154696133,
|
| 350 |
+
"logits/chosen": -2.794644832611084,
|
| 351 |
+
"logits/rejected": -2.7910995483398438,
|
| 352 |
+
"logps/chosen": -58.40827560424805,
|
| 353 |
+
"logps/rejected": -79.7098617553711,
|
| 354 |
+
"loss": 1.2398,
|
| 355 |
+
"rewards/accuracies": 0.25,
|
| 356 |
+
"rewards/chosen": -3.556912660598755,
|
| 357 |
+
"rewards/margins": 1.693927526473999,
|
| 358 |
+
"rewards/rejected": -5.250839710235596,
|
| 359 |
+
"step": 240
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.0,
|
| 363 |
+
"learning_rate": 0.0007237569060773481,
|
| 364 |
+
"logits/chosen": -2.703537702560425,
|
| 365 |
+
"logits/rejected": -2.7038962841033936,
|
| 366 |
+
"logps/chosen": -40.56044006347656,
|
| 367 |
+
"logps/rejected": -43.81157684326172,
|
| 368 |
+
"loss": 1.3514,
|
| 369 |
+
"rewards/accuracies": 0.10000000149011612,
|
| 370 |
+
"rewards/chosen": -2.4380924701690674,
|
| 371 |
+
"rewards/margins": 0.2092890739440918,
|
| 372 |
+
"rewards/rejected": -2.647381544113159,
|
| 373 |
+
"step": 250
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.04,
|
| 377 |
+
"learning_rate": 0.0006961325966850829,
|
| 378 |
+
"logits/chosen": -2.870227098464966,
|
| 379 |
+
"logits/rejected": -2.891268253326416,
|
| 380 |
+
"logps/chosen": -67.85215759277344,
|
| 381 |
+
"logps/rejected": -91.36506652832031,
|
| 382 |
+
"loss": 1.2686,
|
| 383 |
+
"rewards/accuracies": 0.2750000059604645,
|
| 384 |
+
"rewards/chosen": -3.6993489265441895,
|
| 385 |
+
"rewards/margins": 1.7562892436981201,
|
| 386 |
+
"rewards/rejected": -5.455638408660889,
|
| 387 |
+
"step": 260
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.08,
|
| 391 |
+
"learning_rate": 0.0006685082872928176,
|
| 392 |
+
"logits/chosen": -3.1669762134552,
|
| 393 |
+
"logits/rejected": -3.1949028968811035,
|
| 394 |
+
"logps/chosen": -30.239299774169922,
|
| 395 |
+
"logps/rejected": -42.69363021850586,
|
| 396 |
+
"loss": 0.7342,
|
| 397 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 398 |
+
"rewards/chosen": -0.33911675214767456,
|
| 399 |
+
"rewards/margins": 0.5873344540596008,
|
| 400 |
+
"rewards/rejected": -0.9264512062072754,
|
| 401 |
+
"step": 270
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.12,
|
| 405 |
+
"learning_rate": 0.0006408839779005525,
|
| 406 |
+
"logits/chosen": -3.065441131591797,
|
| 407 |
+
"logits/rejected": -3.0655035972595215,
|
| 408 |
+
"logps/chosen": -39.723567962646484,
|
| 409 |
+
"logps/rejected": -53.760826110839844,
|
| 410 |
+
"loss": 1.3162,
|
| 411 |
+
"rewards/accuracies": 0.15000000596046448,
|
| 412 |
+
"rewards/chosen": -2.2118256092071533,
|
| 413 |
+
"rewards/margins": 0.6879772543907166,
|
| 414 |
+
"rewards/rejected": -2.8998026847839355,
|
| 415 |
+
"step": 280
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.16,
|
| 419 |
+
"learning_rate": 0.0006132596685082873,
|
| 420 |
+
"logits/chosen": -2.6939263343811035,
|
| 421 |
+
"logits/rejected": -2.6940252780914307,
|
| 422 |
+
"logps/chosen": -60.013755798339844,
|
| 423 |
+
"logps/rejected": -58.038536071777344,
|
| 424 |
+
"loss": 2.1225,
|
| 425 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 426 |
+
"rewards/chosen": -3.602609634399414,
|
| 427 |
+
"rewards/margins": 0.10532107204198837,
|
| 428 |
+
"rewards/rejected": -3.70793080329895,
|
| 429 |
+
"step": 290
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.2,
|
| 433 |
+
"learning_rate": 0.000585635359116022,
|
| 434 |
+
"logits/chosen": -2.695570468902588,
|
| 435 |
+
"logits/rejected": -2.6920197010040283,
|
| 436 |
+
"logps/chosen": -46.42293930053711,
|
| 437 |
+
"logps/rejected": -44.63296127319336,
|
| 438 |
+
"loss": 0.9542,
|
| 439 |
+
"rewards/accuracies": 0.17499999701976776,
|
| 440 |
+
"rewards/chosen": -2.1268885135650635,
|
| 441 |
+
"rewards/margins": 0.09368989616632462,
|
| 442 |
+
"rewards/rejected": -2.22057843208313,
|
| 443 |
+
"step": 300
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.24,
|
| 447 |
+
"learning_rate": 0.000558011049723757,
|
| 448 |
+
"logits/chosen": -3.1045467853546143,
|
| 449 |
+
"logits/rejected": -3.1222219467163086,
|
| 450 |
+
"logps/chosen": -22.447757720947266,
|
| 451 |
+
"logps/rejected": -27.05214500427246,
|
| 452 |
+
"loss": 0.7383,
|
| 453 |
+
"rewards/accuracies": 0.15000000596046448,
|
| 454 |
+
"rewards/chosen": -0.1806814968585968,
|
| 455 |
+
"rewards/margins": 0.14409010112285614,
|
| 456 |
+
"rewards/rejected": -0.32477161288261414,
|
| 457 |
+
"step": 310
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.28,
|
| 461 |
+
"learning_rate": 0.0005303867403314917,
|
| 462 |
+
"logits/chosen": -3.1156842708587646,
|
| 463 |
+
"logits/rejected": -3.130932569503784,
|
| 464 |
+
"logps/chosen": -21.35702133178711,
|
| 465 |
+
"logps/rejected": -20.4589900970459,
|
| 466 |
+
"loss": 0.7337,
|
| 467 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 468 |
+
"rewards/chosen": -0.29580003023147583,
|
| 469 |
+
"rewards/margins": 0.08602263033390045,
|
| 470 |
+
"rewards/rejected": -0.3818226456642151,
|
| 471 |
+
"step": 320
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.32,
|
| 475 |
+
"learning_rate": 0.0005027624309392266,
|
| 476 |
+
"logits/chosen": -3.050220012664795,
|
| 477 |
+
"logits/rejected": -3.046596050262451,
|
| 478 |
+
"logps/chosen": -24.79575538635254,
|
| 479 |
+
"logps/rejected": -33.02082443237305,
|
| 480 |
+
"loss": 0.6601,
|
| 481 |
+
"rewards/accuracies": 0.25,
|
| 482 |
+
"rewards/chosen": -0.31610769033432007,
|
| 483 |
+
"rewards/margins": 0.4491572976112366,
|
| 484 |
+
"rewards/rejected": -0.7652650475502014,
|
| 485 |
+
"step": 330
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.36,
|
| 489 |
+
"learning_rate": 0.00047513812154696136,
|
| 490 |
+
"logits/chosen": -3.1225409507751465,
|
| 491 |
+
"logits/rejected": -3.142671823501587,
|
| 492 |
+
"logps/chosen": -34.958404541015625,
|
| 493 |
+
"logps/rejected": -37.874855041503906,
|
| 494 |
+
"loss": 0.6053,
|
| 495 |
+
"rewards/accuracies": 0.30000001192092896,
|
| 496 |
+
"rewards/chosen": -0.23902547359466553,
|
| 497 |
+
"rewards/margins": 0.42017507553100586,
|
| 498 |
+
"rewards/rejected": -0.6592004895210266,
|
| 499 |
+
"step": 340
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.4,
|
| 503 |
+
"learning_rate": 0.00044751381215469617,
|
| 504 |
+
"logits/chosen": -3.3166356086730957,
|
| 505 |
+
"logits/rejected": -3.271238327026367,
|
| 506 |
+
"logps/chosen": -41.72985076904297,
|
| 507 |
+
"logps/rejected": -43.41400909423828,
|
| 508 |
+
"loss": 0.7479,
|
| 509 |
+
"rewards/accuracies": 0.30000001192092896,
|
| 510 |
+
"rewards/chosen": -0.3358752131462097,
|
| 511 |
+
"rewards/margins": 0.2891044020652771,
|
| 512 |
+
"rewards/rejected": -0.6249796152114868,
|
| 513 |
+
"step": 350
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.44,
|
| 517 |
+
"learning_rate": 0.0004198895027624309,
|
| 518 |
+
"logits/chosen": -3.2012417316436768,
|
| 519 |
+
"logits/rejected": -3.2214763164520264,
|
| 520 |
+
"logps/chosen": -29.948162078857422,
|
| 521 |
+
"logps/rejected": -34.02383041381836,
|
| 522 |
+
"loss": 0.7083,
|
| 523 |
+
"rewards/accuracies": 0.15000000596046448,
|
| 524 |
+
"rewards/chosen": -0.20552043616771698,
|
| 525 |
+
"rewards/margins": -0.0057243406772613525,
|
| 526 |
+
"rewards/rejected": -0.19979611039161682,
|
| 527 |
+
"step": 360
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.48,
|
| 531 |
+
"learning_rate": 0.00039226519337016573,
|
| 532 |
+
"logits/chosen": -3.217780590057373,
|
| 533 |
+
"logits/rejected": -3.2134361267089844,
|
| 534 |
+
"logps/chosen": -16.843505859375,
|
| 535 |
+
"logps/rejected": -17.496295928955078,
|
| 536 |
+
"loss": 0.763,
|
| 537 |
+
"rewards/accuracies": 0.10000000149011612,
|
| 538 |
+
"rewards/chosen": -0.05251544713973999,
|
| 539 |
+
"rewards/margins": -0.043300654739141464,
|
| 540 |
+
"rewards/rejected": -0.009214771911501884,
|
| 541 |
+
"step": 370
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.52,
|
| 545 |
+
"learning_rate": 0.0003646408839779006,
|
| 546 |
+
"logits/chosen": -3.168470859527588,
|
| 547 |
+
"logits/rejected": -3.2095096111297607,
|
| 548 |
+
"logps/chosen": -31.317157745361328,
|
| 549 |
+
"logps/rejected": -40.99824523925781,
|
| 550 |
+
"loss": 0.6099,
|
| 551 |
+
"rewards/accuracies": 0.3499999940395355,
|
| 552 |
+
"rewards/chosen": -0.06307787448167801,
|
| 553 |
+
"rewards/margins": 0.3778603971004486,
|
| 554 |
+
"rewards/rejected": -0.4409382939338684,
|
| 555 |
+
"step": 380
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.56,
|
| 559 |
+
"learning_rate": 0.0003370165745856354,
|
| 560 |
+
"logits/chosen": -3.1807992458343506,
|
| 561 |
+
"logits/rejected": -3.1729633808135986,
|
| 562 |
+
"logps/chosen": -25.59493064880371,
|
| 563 |
+
"logps/rejected": -24.889175415039062,
|
| 564 |
+
"loss": 0.6768,
|
| 565 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 566 |
+
"rewards/chosen": -0.03841208666563034,
|
| 567 |
+
"rewards/margins": 0.2406325787305832,
|
| 568 |
+
"rewards/rejected": -0.2790446877479553,
|
| 569 |
+
"step": 390
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.6,
|
| 573 |
+
"learning_rate": 0.00030939226519337016,
|
| 574 |
+
"logits/chosen": -3.1939032077789307,
|
| 575 |
+
"logits/rejected": -3.1782355308532715,
|
| 576 |
+
"logps/chosen": -14.640347480773926,
|
| 577 |
+
"logps/rejected": -13.197868347167969,
|
| 578 |
+
"loss": 0.6587,
|
| 579 |
+
"rewards/accuracies": 0.15000000596046448,
|
| 580 |
+
"rewards/chosen": -0.013051311485469341,
|
| 581 |
+
"rewards/margins": 0.11761553585529327,
|
| 582 |
+
"rewards/rejected": -0.1306668370962143,
|
| 583 |
+
"step": 400
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.64,
|
| 587 |
+
"learning_rate": 0.00028176795580110497,
|
| 588 |
+
"logits/chosen": -3.2121384143829346,
|
| 589 |
+
"logits/rejected": -3.242738723754883,
|
| 590 |
+
"logps/chosen": -26.492828369140625,
|
| 591 |
+
"logps/rejected": -33.63092041015625,
|
| 592 |
+
"loss": 0.6547,
|
| 593 |
+
"rewards/accuracies": 0.2750000059604645,
|
| 594 |
+
"rewards/chosen": -0.15410800278186798,
|
| 595 |
+
"rewards/margins": 0.2064083367586136,
|
| 596 |
+
"rewards/rejected": -0.36051633954048157,
|
| 597 |
+
"step": 410
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.68,
|
| 601 |
+
"learning_rate": 0.0002541436464088398,
|
| 602 |
+
"logits/chosen": -3.217862367630005,
|
| 603 |
+
"logits/rejected": -3.2220897674560547,
|
| 604 |
+
"logps/chosen": -45.029319763183594,
|
| 605 |
+
"logps/rejected": -50.58649444580078,
|
| 606 |
+
"loss": 0.7074,
|
| 607 |
+
"rewards/accuracies": 0.30000001192092896,
|
| 608 |
+
"rewards/chosen": -0.21466748416423798,
|
| 609 |
+
"rewards/margins": 0.43003931641578674,
|
| 610 |
+
"rewards/rejected": -0.644706666469574,
|
| 611 |
+
"step": 420
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.72,
|
| 615 |
+
"learning_rate": 0.0002265193370165746,
|
| 616 |
+
"logits/chosen": -3.2857704162597656,
|
| 617 |
+
"logits/rejected": -3.285273313522339,
|
| 618 |
+
"logps/chosen": -24.04534912109375,
|
| 619 |
+
"logps/rejected": -25.24020767211914,
|
| 620 |
+
"loss": 0.7055,
|
| 621 |
+
"rewards/accuracies": 0.22499999403953552,
|
| 622 |
+
"rewards/chosen": -0.05270111560821533,
|
| 623 |
+
"rewards/margins": 0.19839158654212952,
|
| 624 |
+
"rewards/rejected": -0.25109270215034485,
|
| 625 |
+
"step": 430
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.76,
|
| 629 |
+
"learning_rate": 0.0001988950276243094,
|
| 630 |
+
"logits/chosen": -3.1307530403137207,
|
| 631 |
+
"logits/rejected": -3.097612142562866,
|
| 632 |
+
"logps/chosen": -25.230710983276367,
|
| 633 |
+
"logps/rejected": -27.142175674438477,
|
| 634 |
+
"loss": 0.6265,
|
| 635 |
+
"rewards/accuracies": 0.2750000059604645,
|
| 636 |
+
"rewards/chosen": 0.06528893858194351,
|
| 637 |
+
"rewards/margins": 0.3360690474510193,
|
| 638 |
+
"rewards/rejected": -0.27078011631965637,
|
| 639 |
+
"step": 440
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.8,
|
| 643 |
+
"learning_rate": 0.0001712707182320442,
|
| 644 |
+
"logits/chosen": -3.210901975631714,
|
| 645 |
+
"logits/rejected": -3.250978469848633,
|
| 646 |
+
"logps/chosen": -16.752582550048828,
|
| 647 |
+
"logps/rejected": -30.28323745727539,
|
| 648 |
+
"loss": 0.5541,
|
| 649 |
+
"rewards/accuracies": 0.30000001192092896,
|
| 650 |
+
"rewards/chosen": 0.05518122762441635,
|
| 651 |
+
"rewards/margins": 0.479708731174469,
|
| 652 |
+
"rewards/rejected": -0.42452749609947205,
|
| 653 |
+
"step": 450
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.84,
|
| 657 |
+
"learning_rate": 0.000143646408839779,
|
| 658 |
+
"logits/chosen": -3.048260450363159,
|
| 659 |
+
"logits/rejected": -3.1042888164520264,
|
| 660 |
+
"logps/chosen": -33.8494758605957,
|
| 661 |
+
"logps/rejected": -49.50830841064453,
|
| 662 |
+
"loss": 0.5998,
|
| 663 |
+
"rewards/accuracies": 0.3499999940395355,
|
| 664 |
+
"rewards/chosen": 0.0009274661424569786,
|
| 665 |
+
"rewards/margins": 0.6367529630661011,
|
| 666 |
+
"rewards/rejected": -0.635825514793396,
|
| 667 |
+
"step": 460
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.88,
|
| 671 |
+
"learning_rate": 0.0001160220994475138,
|
| 672 |
+
"logits/chosen": -3.272062301635742,
|
| 673 |
+
"logits/rejected": -3.2841033935546875,
|
| 674 |
+
"logps/chosen": -16.145034790039062,
|
| 675 |
+
"logps/rejected": -19.171789169311523,
|
| 676 |
+
"loss": 0.6134,
|
| 677 |
+
"rewards/accuracies": 0.22499999403953552,
|
| 678 |
+
"rewards/chosen": 0.10014114528894424,
|
| 679 |
+
"rewards/margins": 0.250404953956604,
|
| 680 |
+
"rewards/rejected": -0.15026383101940155,
|
| 681 |
+
"step": 470
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.92,
|
| 685 |
+
"learning_rate": 8.839779005524861e-05,
|
| 686 |
+
"logits/chosen": -3.1270499229431152,
|
| 687 |
+
"logits/rejected": -3.1859707832336426,
|
| 688 |
+
"logps/chosen": -28.680988311767578,
|
| 689 |
+
"logps/rejected": -39.859764099121094,
|
| 690 |
+
"loss": 0.6342,
|
| 691 |
+
"rewards/accuracies": 0.32499998807907104,
|
| 692 |
+
"rewards/chosen": -0.053285278379917145,
|
| 693 |
+
"rewards/margins": 0.39891186356544495,
|
| 694 |
+
"rewards/rejected": -0.4521971344947815,
|
| 695 |
+
"step": 480
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.96,
|
| 699 |
+
"learning_rate": 6.0773480662983424e-05,
|
| 700 |
+
"logits/chosen": -3.140641212463379,
|
| 701 |
+
"logits/rejected": -3.090985059738159,
|
| 702 |
+
"logps/chosen": -32.661651611328125,
|
| 703 |
+
"logps/rejected": -32.10502624511719,
|
| 704 |
+
"loss": 0.7772,
|
| 705 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 706 |
+
"rewards/chosen": -0.3610732853412628,
|
| 707 |
+
"rewards/margins": 0.07261800020933151,
|
| 708 |
+
"rewards/rejected": -0.43369120359420776,
|
| 709 |
+
"step": 490
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.0,
|
| 713 |
+
"learning_rate": 3.3149171270718233e-05,
|
| 714 |
+
"logits/chosen": -3.1711134910583496,
|
| 715 |
+
"logits/rejected": -3.190059185028076,
|
| 716 |
+
"logps/chosen": -36.2880973815918,
|
| 717 |
+
"logps/rejected": -48.17897415161133,
|
| 718 |
+
"loss": 0.613,
|
| 719 |
+
"rewards/accuracies": 0.375,
|
| 720 |
+
"rewards/chosen": -0.25595012307167053,
|
| 721 |
+
"rewards/margins": 0.6426823139190674,
|
| 722 |
+
"rewards/rejected": -0.8986324071884155,
|
| 723 |
+
"step": 500
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.0,
|
| 727 |
+
"eval_logits/chosen": -3.225074052810669,
|
| 728 |
+
"eval_logits/rejected": -3.2351696491241455,
|
| 729 |
+
"eval_logps/chosen": -28.331165313720703,
|
| 730 |
+
"eval_logps/rejected": -31.33060073852539,
|
| 731 |
+
"eval_loss": 0.7142000794410706,
|
| 732 |
+
"eval_rewards/accuracies": 0.22200000286102295,
|
| 733 |
+
"eval_rewards/chosen": -0.15271247923374176,
|
| 734 |
+
"eval_rewards/margins": 0.16096290946006775,
|
| 735 |
+
"eval_rewards/rejected": -0.3136754035949707,
|
| 736 |
+
"eval_runtime": 411.5707,
|
| 737 |
+
"eval_samples_per_second": 2.43,
|
| 738 |
+
"eval_steps_per_second": 0.304,
|
| 739 |
+
"step": 500
|
| 740 |
+
}
|
| 741 |
+
],
|
| 742 |
+
"logging_steps": 10,
|
| 743 |
+
"max_steps": 512,
|
| 744 |
+
"num_input_tokens_seen": 0,
|
| 745 |
+
"num_train_epochs": 3,
|
| 746 |
+
"save_steps": 500,
|
| 747 |
+
"total_flos": 0.0,
|
| 748 |
+
"train_batch_size": 4,
|
| 749 |
+
"trial_name": null,
|
| 750 |
+
"trial_params": null
|
| 751 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:42c06184f83dd8a6877e02e782c90e9afb80bdbc086da2fed33b592b22abde6d
|
| 3 |
+
size 4664
|